Gosper's Algorithm for Life

[Paul Dietz]

I would be grateful if someone could send me a copy of an
implementation of Gosper's algorithm for 2-D cellular automata
(Physica 10D (1984), pp 75-80), preferably in C or Pascal. I am
thinking about using it in a data structures course; it is a nice
example of short-circuiting function evaluation by storing previously
computed results in a hash table.

I'd also be interested in hearing about practical experience with
his algorithm, and comparisons to other algorithms.

Paul F. Dietz
di...@cs.rochester.edu

[Derek Bell]

In article <1989Dec27.1...@cs.rochester.edu> di...@cs.rochester.edu (Paul Dietz) writes:
}I would be grateful if someone could send me a copy of an
}implementation of Gosper's algorithm for 2-D cellular automata
}(Physica 10D (1984), pp 75-80), preferably in C or Pascal.
}

}I'd also be interested in hearing about practical experience with
}his algorithm, and comparisons to other algorithms.

I'm interested too.

--
Derek Bell |To everyone is given the key to heaven, the same key opens
db...@maths.tcd.ie | the gates of hell -- Ancient proverb
be...@vax1.tcd.ie |

[Paul Dietz]

Thanks for the folks who sent me a copy of Lucas' C implementation
of Gosper's Hashlife algorithm.

I was wondering how good you could do with other algorithms. I came
up with an implementation of an algorithm on my Sparcstation 1 that
does all 1103 generations of the R-pentomino, including the gliders,
in 1.3 seconds (no display), and runs

o.....o.
..o...o.
..o..o.o
.o.o....

(a pattern by Wainright recommended to me by Gosper) out to 10,000
generations in 45 seconds (again, without display).

Paul